SEARCH

SEARCH BY CITATION

Keywords:

  • channelrhodopsin-2;
  • endoscope;
  • fluorescent imaging;
  • multielectrode;
  • optogenetics

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Controlling neural activity with high spatio-temporal resolution is desired for studying how neural circuit dynamics control animal behavior. Conventional methods for manipulating neural activity, such as electrical microstimulation or pharmacological blockade, have poor spatial and/or temporal resolution. Algal protein channelrhodopsin-2 (ChR2) enables millisecond-precision control of neural activity. However, a photostimulation method for high spatial resolution mapping in vivo is yet to be established. Here, we report a novel optical/electrical probe, consisting of optical fiber bundles and metal electrodes. Optical fiber bundles were used as a brain-insertable endoscope for image transfer and stimulating light delivery. Light-induced activity from ChR2-expressing neurons was detected with electrodes bundled to the endoscope, enabling verification of light-evoked action potentials. Photostimulation through optical fiber bundles of transgenic mice expressing ChR2 in layer 5 cortical neurons resulted in single-whisker movement, indicating spatially restricted activation of neurons in vivo. The probe system described here and a combination of various photoactive molecules will facilitate studies on the causal link between specific neural activity patterns and behavior.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

A fundamental problem in neuroscience is how spatially and temporally complex patterns of neural activity mediate higher brain functions, such as specific actions and perceptions. To answer this question, not only recording, but also controlling neural activity with high spatio-temporal resolution is required. Electrical stimulation has long been used to investigate neural substrates for a number of motor and cognitive functions (Fritsch & Hitzig, 1870; Penfield & Boldrey, 1937; Asanuma et al., 1968; Salzman et al., 1990). However, this method has some shortcomings – the inability to selectively target neuronal subtypes, limited spatial resolution with extracellular stimulation, and the limited number of neurons (typically one cell) that can be activated with intracellular stimulation.

Recently, light-sensitive cation channels such as algal protein channelrhodopsin-2 (ChR2) have been adopted to stimulate neurons by light. This method offers many advantages over conventional methods for controlling neural activity, such as millisecond-precision, lack of toxicity and genetic control of target cell types (Boyden et al., 2005; Ishizuka et al., 2006). Combination of cell type-specific expression of ChR2 and photostimulation revealed particular roles of various types of neurons (Adamantidis et al., 2007; Cardin et al., 2009; Tsai et al., 2009). Light-induced silencing of neural activity is also possible using a light-driven chloride pump, such as halorhodopsin (Han & Boyden, 2007; Zhang et al., 2007). However, controlling neural activity in living animals by light with high spatial resolution is yet to be achieved.

To apply this photic control method of neural activity in vivo, a combined probe consisted of optical fiber and electrode is implanted in the brain to stimulate and record neural activity. A single sharp electrode combined with an optical fiber called ‘optrode’ has been used to deliver stimulating light and to monitor neural activity (Fig. 1A; Gradinaru et al., 2007). Because light propagates bidirectionally through optical fiber, the optical fiber for stimulating light delivery can also be used for fluorescence detection (LeChasseur et al., 2011). For high-throughput neural activity recording in vivo, the multi-channel version of optrode, which consists of single optical fiber and multi-channel electrodes, has recently been reported (Fig. 1B; Zhang et al., 2009; Royer et al., 2010; Anikeeva et al., 2012). These types of probes enable us to control and record activity of multiple neurons. However, these probes are not suited for light stimulation with high spatial resolution, because only one optical channel is equipped. To control multiple neural activity independently, multiple optical channels should be required.

image

Figure 1.  Various types of optical/electrical probes for deep brain. (A) An optical/electrical probe having single optical fiber and single electrode. (B) Three types of single optical fiber–multiple electrode combination. Left – wire-wound tetrodes are combined with an optical fiber. Center –‘Michigan probe’ is integrated with an optical fiber. Right –‘Utah’ multi-electrode array is combined with a tapered optical fiber. The optical fiber is Au-coated, and also works as an electrode. (C) A combination of multiple optical fiber and multiple electrode. Multiple optical fiber cores are bundled with a spacing of 3.3 μm in a single optical fiber bundle.

Download figure to PowerPoint

Brain-insertable microendoscope has been used to visualize deep brain regions (Jung et al., 2004; Vincent et al., 2006). Optical probes used in these studies were made of a gradient refractive index lens or optical fiber bundles, and their outer diameters were typically 0.25–1 mm for minimally invasive insertion into the solid tissue. With these types of endoscopes, in vivo imaging of fluorescent-labeled cells and neuronal activity measurement with calcium-sensitive dyes were reported (Jung et al., 2004; Vincent et al., 2006). In principle, these microendoscopes can also be used for delivering stimulating light, but such application has not been reported so far.

We report here a new method for controlling neural activity with high spatio-temporal resolution, which consists of optical fiber bundle-based endoscopes and metal microelectrodes (Fig. 1C). This probe enables targeted photostimulation with high spatial resolution, while monitoring light-evoked neural activity. Using this optical fiber bundle-based endoscope, we first show that this new probe is useful for stimulating neurons with high resolution in living animals. We then show that photostimulation of the primary motor cortex of transgenic mice expressing ChR2 in layer 5 cortical neurons can evoke single-whisker movement, indicating spatially restricted activation of neurons in deep brain regions.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Plasmids

DNA encoding ChR2-enhanced yellow fluorescent protein (EYFP; a gift from K. Deisseroth), enhanced green fluorescent protein (EGFP) and tdTomato were subcloned into the pCAGGS expression vector (a gift from Jun-ichi Miyazaki, Osaka University, Osaka, Japan).

Animals

Photostimulation and electrophysiological recording experiments were performed on ICR mice (20–32 g, aged 4–12 weeks) that were anesthetized by a ketamine and xylazine mixture (90 mg/kg ketamine, 5 mg/kg xylazine). For whisker movement experiments, Thy1-ChR2-EYFP transgenic mice [Jackson Laboratory strain B6.Cg-Tg(Thy1-COP4/EYFP)18Gfng/J; Arenkiel et al., 2007] were used (20–30 g, aged 6–12 weeks). Animals were given food and water ad libitum. The protocols used for all animal experiments in this study were approved by the animal research committee of Osaka Bioscience Institute.

In utero electroporation

E15.5 timed-pregnant ICR mice were deeply anesthetized with Nembutal (50 mg/kg) in saline. A midline laparotomy was performed to expose the uterus. For DNA microinjection, glass capillary tubes (GC150TF-10; Harvard Apparatus, Holliston, MA, USA) were pulled using a micropipette puller (PB-7; Narishige, Tokyo, Japan). The DNA solution contained a mixture of plasmids encoding ChR2-EYFP and EGFP in an 1 : 1 volume ratio, at a final concentration of 2 mg/mL. Approximately 1 μL of DNA solution colored with trypan blue was injected into the lateral ventricle of embryos, and square electric pulses (50 V, 50 ms) were delivered five times at the rate of 1 pulse/s by an electroporator (CUY21EDIT; NepaGene, Chiba, Japan). After electroporation, the uterus was repositioned, and the abdominal wall and skin were sutured. For brain slice recording, tdTomato was used instead of EGFP.

In vivo photostimulation and electrophysiological recording

Custom optical/electrical microprobes (Fibertech, Tokyo, Japan) were used to acquire fluorescent images of brain tissue, to guide stimulating light and to detect neural activity. The probe consisted of three optical fiber bundles (Fujikura, Tokyo, Japan) and 10 tungsten microwires (California Fine Wire, Fremont, CA, USA). These optical fibers and microwires are inserted into a stainless steel tube. A confocal scanner unit (FV300; Olympus, Tokyo, Japan) was used to visualize fluorescent images and to scan stimulating light. EGFP was excited with 473-nm solid-state laser (CNI, Changchun, China), and emitted fluorescence (495–540 nm) was detected with a GaAsP photomultiplier unit (H7422PA-40; Hamamatsu Photonics K.K., Shizuoka, Japan). Stimulating light was coupled into the optical fiber bundles with an objective lens (MPlan N 20 × /0.4 NA, Olympus). The coupling efficiency between the objective lens and a single core of the optical fiber bundle was ∼10–20% for the 473-nm laser. ChR2 was also excited with the 473-nm laser. An acousto-optical tunable filter (AOTFnC-400.650; AA Optoelectronic, Orsay, France) was used for controlling the intensity of the laser beam. The whole system was controlled with custom software written in labview 7.1 (National Instruments, Dallas, TX, USA). Neural waveforms were amplified (× 2000) and filtered (300–5000 Hz) with a multichannel amplifier (Model 3600; A-M systems, Sequim, WA, USA). Amplified signals were digitized with an analog-to-digital converter board (PCI-6259; National Instruments). The sampling frequency was 20 kHz, and the signal was digitally high-pass filtered at an 800-Hz cutoff. Electrophysiological data were processed and analysed with matlab 2006b (Mathworks, Natick, MA, USA).

Motor cortex stimulation

A 215-μm-diameter optical fiber bundle (FIGH-03-215S, Fujikura) was used as an endoscope for stimulating light delivery. The tip was formed into a cone shape and was inserted into the primary motor cortex (1.5 mm anterior, 1.0 mm lateral from bregma, 0.5 mm deep from brain surface) of the anesthetized mouse. Initially, brief light pulses of several different light intensities (0.06, 0.3, 1.5 and 6 mW at endoscope tip) were used to determine whether any movement was evoked. If movement was detected at a certain light intensity, a light stimulation series (20 steps of light intensity) was applied. Light intensity was increased by 1.1 × at one step, and the stimuli were delivered in ascending order. At each step, light stimulation contained five 40-ms light pulses with 500-ms intervals. Whisker movements were captured at 50 frames/s with a video camera (RM-6740CL; JAI, Copenhagen, Denmark). We classified trials as ‘single-whisker movement’, where only one whisker was diffracted or a large (twice) difference was detected between the best and second-best whisker in movement amplitude at threshold. Video images were analysed using ImageJ (http://rsb.info.nih.gov/ij/) and matlab.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

An optical/electrical probe system

We describe here a method for ChR2-assisted optical control of neural activity in vivo with high spatio-temporal resolution. A newly designed optical/electrical probe was used to image neurons, deliver stimulating light with high spatial resolution, and record neural activity in living animals (Fig. 2A). The device was composed of three optical fiber bundles (80 or 125 μm diameter) and 10 tungsten microelectrodes (Fig. 2B; Table 1). The probe tip had a 45 º beveled edge for minimizing brain damage. Smaller diameter electrodes (7.6 μm diameter) were gold-plated to reduce electrical impedance. The optical fiber bundle, which consisted of hundreds of optical fibers, transmitted an image to a remote end (Fig. 2C). Because light propagates bidirectionally in the optical fibers, the bundle could deliver illuminating light to the neural tissue and transmit fluorescent images back to the photodetector (Fig. 2A). Each optical fiber bundle consisted of 1.9-μm-diameter single-mode optical fibers, and the spacing of each fiber was 3.3 μm, which determined the spatial resolution of a transferred image. The numerical aperture of each fiber is 0.41, and the half angle of emission from the fiber in water was approximately 10 º (Fig. 2D). A previous study showed that the spatial resolution of an optical fiber bundle-based endoscope is sufficient to visualize fluorescently labeled neurons at single-cell resolution (Vincent et al., 2006). Stimulating light was deflected by a pair of galvanometer scanners (Fig. 2A), enabling stimulating light to be sent to a single fiber core in the optical fiber bundles (Fig. 2D). This feature is important for controlling neural activity with high spatial resolution (see below).

image

Figure 2.  An optical/electrical probe and channelrhodopsin-2 (ChR2) expression in the brain tissue. (A) Schematic of the optical/electrical probe system. Optical fiber bundles were used to send stimulating light and to transfer a fluorescent image. Neuronal activities were detected with metal electrodes bundled with the optical fiber bundles. Excitation light was deflected by a pair of galvanometer scanners and was coupled into the optical fiber bundles with an objective lens (× 20, 0.4 NA). (B) Tip of the probe: a, optical fiber bundles; b, tungsten wire; c, stainless steel tubing. (C) Microscopic image of the surface of the probe, showing an array of optical fiber cores. (D) Light irradiation from one optical fiber core of the probe in fluorescein sodium solution. (E) Fluorescent image of cortical neurons. A ChR2-EYFP/EGFP co-electroporated mouse was fixed on the stage of a fluorescent stereoscope, and a fluorescent image was taken through a hole in the skull. (F) Coronal section of the cerebral cortex showing layer 2/3-restricted expression of EGFP (yellow, EGFP; red, DNA label). (G) Fluorescent image acquired with the endoscope inserted into the cerebral cortex of the ChR2-EYFP/EGFP co-electroporated mouse. Three white dashed circles indicate the field of view of the endoscope. Red arrows indicate EGFP fluorescence from cortical neurons. (H) The cerebral cortex was electroporated with a mixture of tdTomato and ChR2-EYFP expression plasmid. A representative confocal image of postnatal day 14 cortical slice from the electroporated hemisphere was shown. About 20% of tdTomato-labeled neurons (red) were strongly expressing ChR2-EYFP (green). EYFP, enhanced yellow fluorescent protein.

Download figure to PowerPoint

Table 1. Characteristics of the optical/electrical probe
Probe125 × 380 × 3
Probe diameter (μm)400300
Optical fiber bundle diameter (μm)12580
Image circle diameter (μm)11576
Number of fibers907410
Numerical aperture0.410.41
Electrode diameter (μm)12.77.6
Electrode materialTungstenGold-plated tungsten
Electrode impedance at 1 kHz (kΩ)∼800∼300

We used an in utero electroporation technique for targeted expression of ChR2 to projection neurons in layer 2/3 of the mouse cerebral cortex. Because the fluorescence intensity of ChR2-EYFP was too weak to visualize ChR2-expressing cells through the optical fiber bundle, EGFP were co-electroporated with ChR2-EYFP as a marker of transfected cell. Previous work has shown that multiple plasmids can be introduced into the same cells by in utero electroporation (Saito & Nakatsuji, 2001; Mizuno et al., 2007). First, we confirmed that roughly 50% of layer 2/3 projection neurons were labeled with EGFP (Fig. 2E and F), we then evaluated the co-expression rate of ChR2 and fluorescent marker protein. For this purpose, we employed a red fluorescent protein tdTomato instead of EGFP, for separating the fluorescent signal of marker protein from ChR2-EYFP fluorescence. Although ChR2-EYFP fluorescence was detectable in almost all tdTomato-labeled neurons, only about 20% of tdTomato-labeled neurons strongly express ChR2-EYFP (Fig. 2H). This indicates that expression efficiency of ChR2-EYFP was much lower than that of EGFP or tdTomato. Hence, we used EGFP fluorescence as a marker for the ChR2-expressing region, not for individual ChR2-expressing cells.

With the optical/electrical probe inserted into the cerebral cortex of the anesthetized mouse in which the EGFP and ChR2-EYFP gene were transfected into layer 2/3 cortical projection neurons, EGFP-labeled neurons were clearly visualized (Fig. 2G). This layer-restricted expression pattern of ChR2 by in utero electroporation (Fig. 2F and H) is suited for restricting the region of photoactivation by our optical fiber bundle-based photostimulation method, because the axial intensity distribution of stimulating light is less localized compared with radial distribution (Fig. 2D).

Electrophysiological recording of neural activity with the probe

We first recorded spontaneous neural activity of cortical neurons with the probe. Spontaneous activity was detected by multiple electrodes in the probe (Fig. 3). In most cases, each electrode detected multiple unit activities (Fig. 3), this is probably because we used low-impedance electrodes (∼300–800 kΩ at 1 kHz) to monitor activity over a large area. This result indicates that considerable numbers of neurons surrounding the probe are viable and excitable.

image

Figure 3.  Recording neural activity with the optical/electrical probe. Spontaneous activity was recorded by the probe inserted at layer 2/3 of the cerebral cortex. Spike-like activity was detected in five out of 10 electrodes (Nos 5–9) in the probe. Spike overlays were shown in the right side. Inset – electrode positions on the probe tip are displayed.

Download figure to PowerPoint

Optical stimulation and electrophysiological recording with the probe

We then stimulated ChR2-EGFP co-expressing cortical pyramidal neurons in the anesthetized mouse with blue light (473 nm) through the probe. As shown in Fig. 4A, stimulating light was raster-scanned in rectangular areas in the endoscopic field of view. Light-evoked neural activities were recorded with the electrodes bundled with the probe (Fig. 4B). Photostimulation through the probe sometimes evoked both spiking and non-spiking activities. Therefore, in this case, neural waveforms were high-pass filtered to extract action potential-like activity (Fig. 4C). Typical waveforms of light-evoked activity are shown in Fig. 4B. When the site A was stimulated, light-evoked spiking activity was detected at only electrode 1. On the other hand, activity was detected at electrode 2 when stimulating site B (Fig. 4B). No activity was detected with the other eight electrodes in the probe when stimulating either site A or B (data not shown). These results indicate that, using this optical/electrical probe system, stimulating different sites in the endoscopic field of view could evoke different spatial activity patterns of neurons. Sometimes light-evoked activity was detected with two electrodes simultaneously (Fig. 4D and E) but, in most cases, only one electrode in the probe detected light-evoked activity. This is probably due to the relatively large distance between adjacent electrodes in the probe (at least 40 μm apart).

image

Figure 4. In vivo photostimulation and recording neural activity. (A) Stimulating light was irradiated at two sites (A and B) in the endoscopic field of view (green rectangles). One of three optical fiber bundles in the probe is displayed. White dots indicate electrode position. Stimulating light was scanned along the zig-zag line in the stimulating area (inset, black line). Scan speed – 8 ms/line; light intensity – 0.7 mW at the probe tip. (B) Neural waveforms detected with electrodes on the probe during light stimulation at sites A and B. Superimposed action potential waveforms are shown on the right-hand side. (C) Wideband signal (‘raw’) and digitally high-pass filtered (800 Hz cutoff) waveform (‘filtered’) recorded from the probe. Extracted spike-like waveforms from the ‘filtered’ trace are shown on the right-hand side. Light intensity was 0.7 mW at the probe tip. Scanning mirror positions are displayed as ‘Galvo_X’ and ‘Galvo_Y’. (D) Light-evoked neural activity was detected by two adjacent electrodes. Stimulating light was irradiated at six sites (a–f) in the endoscopic field of view (white rectangles). One of three optical fiber bundles in the probe is displayed. White dots indicate electrode position. Scan speed – 8 ms/line; light intensity – 0.7 mW at the probe tip. (E) Neural waveforms detected with electrodes on the probe. Wideband signals (non-digitally filtered) are displayed.

Download figure to PowerPoint

To test the spatial resolution of our photostimulation method further, we stimulated various areas in the endoscopic field of view and recorded neural activity from the electrodes. Neural activity-generating points in the endoscopic field of view are shown as small dots in Fig. 5. The dots are color-coded according to the electrodes by which spikes were detected. In this experiment, light-induced activities were detected at seven of the 10 electrodes, and only one electrode detected light-induced spiking activity at each stimulation point. This result indicates that our method can activate spatially restricted neuronal populations, and also indicates that by stimulating different positions in the field of view, different sets of neurons can be activated.

image

Figure 5.  Map of spike-generation points in the endoscopic field of view. Large colored dots represent electrode position. Small dots represent spike-generation points. Small dots are color-coded according to the electrodes by which spikes were detected. Scan speed was 8 ms/line and light intensity was 2.0 mW at the probe tip. Superimposed spike waveforms detected with each electrode are shown on both sides. Vertical bars indicate 100 μV, horizontal bars indicate 1 ms.

Download figure to PowerPoint

We next studied the relationship between light intensity and light-induced neural activity. As the intensity of stimulating light increased, the amplitude of neural activity increased (Fig. 6A and B). This result suggests that multiple neurons were activated with high-intensity photostimulation. On the other hand, at minimal light intensity of neural activity generation (0.16 mW), single-unit-like activity was detected (Fig. 6B). Repeated minimal-intensity photostimulation reliably produced single-unit-like activity (Fig. 6D). This activity was specifically evoked when stimulating via the specific fiber core in the stimulating site (Fig. 6C and D). In contrast, stimulating via the other two adjacent fiber cores in the stimulation area (Fig. 6C and D) did not evoke neural activity. Moreover, photostimulation at half the scan speed (32 ms/line; Fig. 6D, right) also evoked spiking activities whose shape was similar to that evoked by normal scan speed (16 ms/line; Fig. 6D, left and center). These observations suggest that the light-evoked spiking activities represent action potential generation rather than subthreshold membrane potential fluctuations. Most of the spiking activity elicited by photostimulation was blocked with tetrodotoxin treatment (Fig. S1). This result also indicates that the recorded activity represents action potential.

image

Figure 6.  Neural activities evoked by various light intensities. (A) Neural waveform during light stimulation of various intensities. Blue bars indicate the light stimulation period. (B) Superimposed spike waveforms of light-evoked activity. (C) Repeated photostimulation of minimal intensity (0.16 mW). The stimulating area (green rectangle) is shown on the endoscopic image (left), and the trajectory of stimulating light in the stimulating area is shown as black lines (right). (D) Neural waveforms during repeated photostimulation of minimal intensity. Blue bars indicate the light stimulation period. Stimulating light position was displayed as ‘Galvo_X’ (X-axis) and ‘Galvo_Y’ (Y-axis). Superimposed spike waveforms are shown as ‘Overlay’. Spike-generating points (red dots) are shown in the stimulating area (green rectangles). Scanning rates of the stimulating light were 16 ms/line (left and center) and 32 ms/line (right) at the X-axis.

Download figure to PowerPoint

In order to precisely estimate the spatial specificity of photostimulation, we measured light-induced action potential generation of ChR2-expressing cells in brain slice preparation. The relationship between light intensity and the distance of photostimulation point from recorded cell was measured (Fig. S2). ChR2-expressing neurons were recorded while the stimulation point (the tip of the optical fiber bundle) was moved in 10-μm steps along the axial axis of the optical fiber bundle, or along a line perpendicular to the bundle’s axial axis. The relationship between stimulating light intensity and probability of light-dependent action potential generation was measured by whole-cell patch-clamp or cell-attached recording (Fig. S2C). When the stimulation point was moved along the axial axis of the optical fiber bundle, the threshold light intensity was unchanged, nevertheless increasing the distance between the recorded cell and stimulation point (Fig. S3A). On the other hand, the threshold light intensity was monotonically increased when the stimulation point was moved along a line perpendicular to the bundle’s axial axis (Fig. S3B). As shown in Fig. S3B, 10–20 μm of horizontal displacement of the stimulation point from the recorded cell significantly increased the threshold intensity for action potential generation. These results indicate that the spatial specificity of this photostimulation method is comparable to the soma size of cortical neurons in the plane perpendicular to the axial axis of the fiber bundle, but the specificity for the axial axis is low. This is compatible with the light intensity distribution examined in the fluorescent solution (Fig. 2D). It should be noted that when the stimulation point was moved along the axial axis of the optical fiber bundle, stimulating light propagates in the extracellular solution (Fig. S3A), not in the brain tissue. This might have caused underestimation of the spatial specificity of photostimulation in the axial axis, because blue light is heavily absorbed by brain tissue (Yizhar et al., 2011).

Photostimulation-induced motor activity

Using the endoscope-based method, we next manipulated motor behavior. Previous studies have shown that electrical stimulation or optogenetic stimulation of the rodent vibrissa motor cortex results in whisker deflections (Hall & Lindholm, 1974; Aravanis et al., 2007). Each whisker on a rodent’s face is connected to single intrinsic muscle (Dorfl, 1982), and studies have shown that low-intensity electrical stimulation can evoke single-whisker movement (Hall & Lindholm, 1974; Brecht et al., 2004). Therefore, the vibrissa system provides an appropriate model to test spatial specificity of neural stimulation. We used a strain of mice expressing ChR2 in projection neurons of cerebral cortex layer 5, output cells of the motor cortex (Arenkiel et al., 2007). An optical fiber bundle was inserted into the vibrissa motor cortex, and a brief light pulse train (40 ms duration, 500 ms interval, 5 repetition) was applied through a single core in the center of the fiber bundle (Fig. 7A and B). To quantify whisker deflection, images of contralateral whiskers were captured with a video camera and their movements were tracked (Fig. 7C). Trajectories of whisker movements are shown in Fig. 7D. Threshold light intensity was determined as the minimum light intensity at which detectable movement of whiskers was observed at more than half of the light pulse in a train. In 68 of the 595 stimulation series (eight sites in eight animals), single-whisker movement was observed at threshold light intensity (Fig. 7D, left) and, in other cases, more than two whisker movements were evoked. Stimulation with a higher light intensity evoked movements of multiple whiskers (Fig. 7D, center and right). Previous electrical microstimulation experiments showed that threshold current intensity and number of deflected whiskers were variable (Brecht et al., 2004). We observed similar results in this ChR2-assisted photostimulation. We stimulated various points in the endoscopic field of view (190 μm diameter). However, no significant difference was observed in stimulation-evoked whisker movement (data not shown). This result indicates that spatial specificity of stimulation is at least as good as that of electrical microstimulation, and also indicates that the endoscope-based photostimulation can activate minimum unit of motor behavior.

image

Figure 7.  Photostimulation-evoked whisker movement. (A) Schematic of the experimental setup. The optical fiber bundle was inserted over the layer 5 cortical neurons of an anesthetized mouse. Whisker movements were monitored using an overhead video camera. (B) The tip of the optical fiber bundle was formed into a cone shape (the half angle was 45º). (C) Sample frame from the video image of a head-fixed channelrhodopsin-2 (ChR2)-transgenic mouse. Movement trajectories of whiskers were measured at the black vertical line. (D) Movement of the contralateral whiskers in response to a 40-ms light pulse of various light intensities through the optical fiber bundle. Black horizontal lines indicate trajectories of whiskers. White vertical lines indicate light pulses. Black arrows indicate light-evoked whisker movements. EYFP, enhanced yellow fluorescent protein.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

In this paper we have described a new optical/electrical probe for controlling neural activity with high spatio-temporal resolution. By using a high-density optical fiber bundle combined with galvano-mirror-based scanning method, we demonstrated that multiple neurons in the endoscopic field of view could be activated independently. In vitro and in vivo experiments suggested that the spatial resolution of photostimulation is comparable to the soma size of cortical neurons in the XY plane (Figs 5 and S3). In addition to better spatial resolution control of neural activity, another advantage of our method is that the activation of a neuron can be verified in real-time by observing action potential generation using the electrodes bundled with the probe (Figs 4–6). This means that one can stimulate neurons with minimal light intensity for target cell activation. Therefore, the combination of optical stimulation and electrical activity monitoring helps to maximize spatial resolution of stimulation and to prevent undesirable side-effects of stimulation.

Several methods for delivering stimulating light to small brain regions have been reported. A metal-coated, sharpened optical fiber was used for both light stimulation and electrical recording (Zhang et al., 2009). Another type of combined probe is based on a dual-core optical fiber – an optical core for delivering stimulating light and an electrolyte-filled hollow core for electrophysiological recording (LeChasseur et al., 2011). The optical apertures in these probes are so small (1–10 μm) that stimulation area is comparable to neuron diameter (Zhang et al., 2009; LeChasseur et al., 2011). Because these probes have only one stimulation and recording site, multiple probes should be arrayed for multi-site stimulation and recording. However, the density of arrayed probes is in general far lower than inter-neuron distance in brain tissue. For example, electrode pitch of ‘Utah’ multiple electrode array is 400 μm (Zhang et al., 2009). In contrast, the optical/electrical probe presented here has hundreds of optical fiber cores with an interval of 3.3 μm. This structure enables us to activate different sets of neurons by stimulating different spots within the endoscopic field of view (80 or 125 μm diameter; Figs 4 and 5). Therefore, the optical fiber bundle-based system presented here offers higher spatial resolution photostimulation compared with these arrayed fiber optic devices. Second, multiphoton excitation was shown to generate an action potential of single ChR2-expressing neurons in dispersedly cultured conditions or in brain slice (Rickgauer & Tank, 2009; Andrasfalvy et al., 2010; Papagiakoumou et al., 2010). Multiphoton excitation is restricted to a tiny focal volume (∼1 femtoliter), which is much smaller than the neuronal cell volume (Denk et al., 1990). Therefore, multiphoton excitation, in principle, enables single-cell resolution control of neural activity. These multiphoton excitation-based techniques can be applied under in vivo conditions. However, because of light scattering, it can only access the brain down to approximately 500 μm in depth (Helmchen & Denk, 2002). Thus, one cannot access subcortical regions of the rodent brain using multiphoton excitation. On the other hand, using an endoscope-based imaging system, this depth limitation can be avoided. For example, deeper brain regions, such as the hippocampus (Barretto et al., 2011) or ventral tegmental area (Vincent et al., 2006), can be visualized clearly with an endoscope inserted into the brain. Our endoscope-based imaging/stimulation system is also applicable for controlling neural activity of deep brain structures. Combination of microendoscope and multiphoton excitation (Jung et al., 2004; Barretto et al., 2011) is a good candidate for optical stimulating method with single-cell resolution in the deep brain region. But it seems difficult to integrate multiphoton endoscope with electrodes for neural activity detection, because a lens for concentrating light on the probe tip is needed for multiphoton absorption. Therefore, an optical method for neural activity detection such as calcium imaging is desirable.

We also showed that with the optical fiber bundle-based probe, it is possible to precisely control animal motor behavior. Functional maps of the motor cortex have been constructed on various species using electrical stimulation (Fritsch & Hitzig, 1870; Penfield & Boldrey, 1937; Asanuma, 1975; Brecht et al., 2004). However, the spatial resolution is 0.5–1 mm at best. Recently, transcranial or epidural photostimulation-based motor mapping methods were reported (Ayling et al., 2009; Hira et al., 2009). These methods enable very fast construction of functional maps compared with using microelectrodes; however, because of light scattering the spatial resolution is no better than that of electrical microstimulation-based mapping. Using our optical/electrical probe, high spatial resolution–optical stimulation in the endoscopic field of view is possible. Thus, our method would be useful to reveal the functional microarchitecture of the brain. We did not observe different movement of whiskers when stimulating various points in the endoscopic field of view. The endoscopic field of view may be too small to cover areas for multiple whisker movement, because whisker area occupies a large portion of the rodent primary motor cortex.

In our experiments, fluorescently-labeled neurons (presumably cell bodies) could be observed through the probe (Fig. 2G), but the region where neural activities were detected did not always correspond to the fluorescent signals (Fig. 4A and D). This is primarily due to the fact that ChR2-expressing neurons were only a subset of EGFP-labeled neurons (Fig. 2H). Another possible reason is that the optical fiber bundle-based endoscope has limited spatial resolution and depth of field (Vincent et al., 2006), therefore thin structures such as axons and dendrites are often not visualized. In addition, neurons distant from the endoscope tip cannot be visualized, because the working distance of this optical fiber bundle-based endoscope is nearly zero (Vincent et al., 2006). Stimulating ChR2 located in these non-visualized neurons can also be activated by light (Fig. S3A). This might have caused the poor correlation between fluorescence signal and neural activity generating areas. The widespread subcellular distribution of ChR2 could also interfere with spatially restricted photostimulation. In most neural tissues, long-range axons are intermingled. Thus, targeted photostimulation on the somatodendritic region of a neuron often also excites colocalized axons of distant neurons. Electrical microstimulation suffers the same problem – stimulating current excites both the soma and axon of neurons; therefore, electrical microstimulation activates neurons around the electrode, sometimes as far as millimeters away (Histed et al., 2009). However, in the case of optical stimulation, this problem would be overcome by molecular biological techniques. A recent report has shown that ChR2 fused with the myosin-binding domain from Melanophilin is targeted to the somatodendritic compartment of neurons in vivo (Lewis et al., 2009). This kind of technology for anchoring ChR2 to specific subcellular regions will aid in achieving high spatial resolution when one uses the method presented here or other optical stimulation techniques for controlling neural activity.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

This work was supported by PRESTO (to Y.H.) and CREST (to K.F.), Japan Science and Technology Agency. We thank members of the Nakanishi Laboratory for helpful advice, M. Okazawa for help in preparing plasmid DNAs, H. Mizuno for assistance with in utero electroporation, and T. Yoshida for critical comments on this work. The authors declare that there are no conflicts of interest.

Abbreviations
ChR2

channelrhodopsin-2

EGFP

enhanced green fluorescent protein

EYFP

enhanced yellow fluorescent protein

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Adamantidis, A.R., Zhang, F., Aravanis, A.M., Deisseroth, K. & de Lecea, L. (2007) Neural substrates of awakening probed with optogenetic control of hypocretin neurons. Nature, 450, 420424.
  • Andrasfalvy, B.K., Zemelman, B.V., Tang, J. & Vaziri, A. (2010) Two-photon single-cell optogenetic control of neuronal activity by sculpted light. Proc. Natl. Acad. Sci. USA, 107, 1198111986.
  • Anikeeva, P., Andalman, A.S., Witten, I., Warden, M., Goshen, I., Grosenick, L., Gunaydin, L.A., Frank, L.M. & Deisseroth, K. (2012) Optetrode: a multichannel readout for optogenetic control in freely moving mice. Nat. Neurosci., 15, 163170.
  • Aravanis, A.M., Wang, L.P., Zhang, F., Meltzer, L.A., Mogri, M.Z., Schneider, M.B. & Deisseroth, K. (2007) An optical neural interface: in vivo control of rodent motor cortex with integrated fiberoptic and optogenetic technology. J. Neural Eng., 4, S143S156.
  • Arenkiel, B.R., Peca, J., Davison, I.G., Feliciano, C., Deisseroth, K., Augustine, G.J., Ehlers, M.D. & Feng, G. (2007) In vivo light-induced activation of neural circuitry in transgenic mice expressing channelrhodopsin-2. Neuron, 54, 205218.
  • Asanuma, H. (1975) Recent developments in the study of the columnar arrangement of neurons within the motor cortex. Physiol. Rev., 55, 143156.
  • Asanuma, H., Stoney, S.D. Jr & Abzug, C. (1968) Relationship between afferent input and motor outflow in cat motorsensory cortex. J. Neurophysiol., 31, 670681.
  • Ayling, O.G., Harrison, T.C., Boyd, J.D., Goroshkov, A. & Murphy, T.H. (2009) Automated light-based mapping of motor cortex by photoactivation of channelrhodopsin-2 transgenic mice. Nat. Methods, 6, 219224.
  • Barretto, R.P., Ko, T.H., Jung, J.C., Wang, T.J., Capps, G., Waters, A.C., Ziv, Y., Attardo, A., Recht, L. & Schnitzer, M.J. (2011) Time-lapse imaging of disease progression in deep brain areas using fluorescence microendoscopy. Nat. Med., 17, 223228.
  • Boyden, E.S., Zhang, F., Bamberg, E., Nagel, G. & Deisseroth, K. (2005) Millisecond-timescale, genetically targeted optical control of neural activity. Nat. Neurosci., 8, 12631268.
  • Brecht, M., Krauss, A., Muhammad, S., Sinai-Esfahani, L., Bellanca, S. & Margrie, T.W. (2004) Organization of rat vibrissa motor cortex and adjacent areas according to cytoarchitectonics, microstimulation, and intracellular stimulation of identified cells. J. Comp. Neurol., 479, 360373.
  • Cardin, J.A., Carlen, M., Meletis, K., Knoblich, U., Zhang, F., Deisseroth, K., Tsai, L.H. & Moore, C.I. (2009) Driving fast-spiking cells induces gamma rhythm and controls sensory responses. Nature, 459, 663667.
  • Denk, W., Strickler, J.H. & Webb, W.W. (1990) Two-photon laser scanning fluorescence microscopy. Science, 248, 7376.
  • Dorfl, J. (1982) The musculature of the mystacial vibrissae of the white mouse. J. Anat., 135, 147154.
  • Fritsch, G. & Hitzig, E. (1870) Ueber die elektrishe Erregarkeit des Grosshirns. In Nowinski, W.W. (Ed), The Cerebral Cortex. Thomas, Splingfield, IL, pp. 7396.
  • Gradinaru, V., Thompson, K.R., Zhang, F., Mogri, M., Kay, K., Schneider, M.B. & Deisseroth, K. (2007) Targeting and readout strategies for fast optical neural control in vitro and in vivo. J. Neurosci., 27, 1423114238.
  • Hall, R.D. & Lindholm, E.P. (1974) Organization of motor and somatosensory neocortex in the albino rat. Brain Res., 66, 2338.
  • Han, X. & Boyden, E.S. (2007) Multiple-color optical activation, silencing, and desynchronization of neural activity, with single-spike temporal resolution. PLoS One, 2, e299.
  • Helmchen, F. & Denk, W. (2002) New developments in multiphoton microscopy. Curr. Opin. Neurobiol., 12, 593601.
  • Hira, R., Honkura, N., Noguchi, J., Maruyama, Y., Augustine, G.J., Kasai, H. & Matsuzaki, M. (2009) Transcranial optogenetic stimulation for functional mapping of the motor cortex. J. Neurosci. Methods, 179, 258263.
  • Histed, M.H., Bonin, V. & Reid, R.C. (2009) Direct activation of sparse, distributed populations of cortical neurons by electrical microstimulation. Neuron, 63, 508522.
  • Ishizuka, T., Kakuda, M., Araki, R. & Yawo, H. (2006) Kinetic evaluation of photosensitivity in genetically engineered neurons expressing green algae light-gated channels. Neurosci. Res., 54, 8594.
  • Jung, J.C., Mehta, A.D., Aksay, E., Stepnoski, R. & Schnitzer, M.J. (2004) In vivo mammalian brain imaging using one- and two-photon fluorescence microendoscopy. J. Neurophysiol., 92, 31213133.
  • LeChasseur, Y., Dufour, S., Lavertu, G., Bories, C., Deschenes, M., Vallee, R. & De Koninck, Y. (2011) A microprobe for parallel optical and electrical recordings from single neurons in vivo. Nat. Methods, 8, 319325.
  • Lewis, T.L. Jr, Mao, T., Svoboda, K. & Arnold, D.B. (2009) Myosin-dependent targeting of transmembrane proteins to neuronal dendrites. Nat. Neurosci., 12, 568576.
  • Mizuno, H., Hirano, T. & Tagawa, Y. (2007) Evidence for activity-dependent cortical wiring: formation of interhemispheric connections in neonatal mouse visual cortex requires projection neuron activity. J. Neurosci., 27, 67606770.
  • Papagiakoumou, E., Anselmi, F., Begue, A., de Sars, V., Gluckstad, J., Isacoff, E.Y. & Emiliani, V. (2010) Scanless two-photon excitation of channelrhodopsin-2. Nat. Methods, 7, 848854.
  • Penfield, W. & Boldrey, E. (1937) Somatic motor and sensory representation in the cerebral cortex of man as studied by electrical stimulation. Brain, 60, 389443.
  • Rickgauer, J.P. & Tank, D.W. (2009) Two-photon excitation of channelrhodopsin-2 at saturation. Proc. Natl. Acad. Sci. USA, 106, 1502515030.
  • Royer, S., Zemelman, B.V., Barbic, M., Losonczy, A., Buzsaki, G. & Magee, J.C. (2010) Multi-array silicon probes with integrated optical fibers: light-assisted perturbation and recording of local neural circuits in the behaving animal. Eur. J. Neurosci., 31, 22792291.
  • Saito, T. & Nakatsuji, N. (2001) Efficient gene transfer into the embryonic mouse brain using in vivo electroporation. Dev. Biol., 240, 237246.
  • Salzman, C.D., Britten, K.H. & Newsome, W.T. (1990) Cortical microstimulation influences perceptual judgements of motion direction. Nature, 346, 174177.
  • Tsai, H.C., Zhang, F., Adamantidis, A., Stuber, G.D., Bonci, A., de Lecea, L. & Deisseroth, K. (2009) Phasic firing in dopaminergic neurons is sufficient for behavioral conditioning. Science, 324, 10801084.
  • Vincent, P., Maskos, U., Charvet, I., Bourgeais, L., Stoppini, L., Leresche, N., Changeux, J.P., Lambert, R., Meda, P. & Paupardin-Tritsch, D. (2006) Live imaging of neural structure and function by fibred fluorescence microscopy. EMBO Rep., 7, 11541161.
  • Yizhar, O., Fenno, L.E., Davidson, T.J., Mogri, M. & Deisseroth, K. (2011) Optogenetics in neural systems. Neuron, 71, 934.
  • Zhang, F., Wang, L.P., Brauner, M., Liewald, J.F., Kay, K., Watzke, N., Wood, P.G., Bamberg, E., Nagel, G., Gottschalk, A. & Deisseroth, K. (2007) Multimodal fast optical interrogation of neural circuitry. Nature, 446, 633639.
  • Zhang, J., Laiwalla, F., Kim, J.A., Urabe, H., Van Wagenen, R., Song, Y.K., Connors, B.W., Zhang, F., Deisseroth, K. & Nurmikko, A.V. (2009) Integrated device for optical stimulation and spatiotemporal electrical recording of neural activity in light-sensitized brain tissue. J. Neural Eng., 6, 055007.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Data S1. Materials and Methods.

Fig. S1. Contribution of Na+ cannels to the light dependentspiking activity. (A) Schematic diagram of the experiment. TTX(100 μm, 0.2 μL) was applied to near the probe tipvia a glass pipette. (B) Typical effect of TTX on light elicitedactivity. Light dependent activities were recorded before (Control)and 5 min after drug applications (Saline, TTX). In manycases, light dependent activity was not detected after TTXtreatment (Left). Sometimes transient activity at lightonset was remained after TTX treatment (Right). Laser powerfor stimulation was 0.6 mW.

Fig. S2. Measurement of spatial specificity. (A) Light irradiation at the tip of the optical fiber bundle. Stimulating light was emitted from one core at the tip of the bundle. (B) Upper, Photostimulation of recorded cell with optical fiberbundle. a: Recording pipette, b: Optical fiber bundle.Lower, Stimulating light was emitted at the bundle’stip. (C) Whole-cell current clamp recordings (Upper) orcell-attach recordings (Lower) in response to 0.5 slight pulses of various light intensities. Laser power forphotostimulation was 1.2 mW at maximum light intensity(denoted as 512). Voltage traces during five repetition ofphotostimulation series were displayed. For whole-cell recording,membrane potential at rest was held around −70 mV byinjecting bias current.

Fig. S3. Spatial resolution of action potential generation.Relationships between light intensity and spike probability weremeasured at various photostimulation points. (A) Stimulation pointwas moved along the axial axis of the bundle. Values on the leftside of the graph indicate distance between recorded cell and thetip of the bundle. (B) Stimulation point was moved along a lineperpendicular to the bundle’s axial axis. Values on the leftside of the graph indicate distance between recorded cell and thetip of the bundle. Laser power for photostimulation was 1.2 mWat maximum light intensity (denoted as 512).

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset by Wiley-Blackwell. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
ejn8191_sm_FigS1-S3.pdf3892KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.